Bump to 0.4.0

Allow the user to specify a heading style.

.gitignore

@@ -0,0 +1,7 @@
+*.pyc
+*.egg
+.DS_Store
+/.env
+/dist
+/MANIFEST
+/venv

MANIFEST.in

@@ -0,0 +1 @@
+include README.rst

README.rst

@@ -0,0 +1,41 @@
+Installation
+============
+
+``pip install markdownify``
+
+
+Usage
+=====
+
+Convert some HTML to Markdown:
+
+.. code:: python
+
+    from markdownify import markdownify as md
+    md('<b>Yay</b> <a href="http://github.com">GitHub</a>')  # > '**Yay** [GitHub](http://github.com)'
+
+Specify tags to exclude (blacklist):
+
+.. code:: python
+
+    from markdownify import markdownify as md
+    md('<b>Yay</b> <a href="http://github.com">GitHub</a>', strip=['a'])  # > '**Yay** GitHub'
+
+\...or specify the tags you want to include (whitelist):
+
+.. code:: python
+
+    from markdownify import markdownify as md
+    md('<b>Yay</b> <a href="http://github.com">GitHub</a>', convert=['b'])  # > '**Yay** GitHub'
+
+
+Development
+===========
+
+To run tests:
+
+``python setup.py test``
+
+To lint:
+
+``python setup.py lint``

markdownify/__init__.py

@@ -1,10 +1,19 @@
-from lxml.html.soupparser import fromstring
+from bs4 import BeautifulSoup, NavigableString
 import re
 
 
 convert_heading_re = re.compile(r'convert_h(\d+)')
 line_beginning_re = re.compile(r'^', re.MULTILINE)
 whitespace_re = re.compile(r'[\r\n\s\t ]+')
+FRAGMENT_ID = '__MARKDOWNIFY_WRAPPER__'
+wrapped = '<div id="%s">%%s</div>' % FRAGMENT_ID
+
+
+# Heading styles
+ATX = 'atx'
+ATX_CLOSED = 'atx_closed'
+UNDERLINED = 'underlined'
+SETEXT = UNDERLINED
 
 
 def escape(text):
@@ -13,30 +22,53 @@ def escape(text):
     return text.replace('_', r'\_')
 
 
+def _todict(obj):
+    return dict((k, getattr(obj, k)) for k in dir(obj) if not k.startswith('_'))
+
+
 class MarkdownConverter(object):
-    def __init__(self, tags_to_strip=None, tags_to_convert=None):
-        if tags_to_strip is not None and tags_to_convert is not None:
+    class DefaultOptions:
+        strip = None
+        convert = None
+        autolinks = True
+        heading_style = UNDERLINED
+        bullets = '*+-'  # An iterable of bullet types.
+
+    class Options(DefaultOptions):
+        pass
+
+    def __init__(self, **options):
+        # Create an options dictionary. Use DefaultOptions as a base so that
+        # it doesn't have to be extended.
+        self.options = _todict(self.DefaultOptions)
+        self.options.update(_todict(self.Options))
+        self.options.update(options)
+        if self.options['strip'] is not None and self.options['convert'] is not None:
             raise ValueError('You may specify either tags to strip or tags to'
-                    ' convert, but not both.')
-        self.tags_to_strip = tags_to_strip
-        self.tags_to_convert = tags_to_convert
+                             ' convert, but not both.')
 
     def convert(self, html):
-        soup = fromstring(html)
-        return self.process_tag(soup)
+        # We want to take advantage of the html5 parsing, but we don't actually
+        # want a full document. Therefore, we'll mark our fragment with an id,
+        # create the document, and extract the element with the id.
+        html = wrapped % html
+        soup = BeautifulSoup(html)
+        return self.process_tag(soup.find(id=FRAGMENT_ID), children_only=True)
 
-    def process_tag(self, node):
-        text = self.process_text(node.text)
+    def process_tag(self, node, children_only=False):
+        text = ''
 
         # Convert the children first
-        for el in node.findall('*'):
-            text += self.process_tag(el)
+        for el in node.children:
+            if isinstance(el, NavigableString):
+                text += self.process_text(unicode(el))
+            else:
+                text += self.process_tag(el)
 
-        convert_fn = getattr(self, 'convert_%s' % node.tag, None)
-        if convert_fn and self.should_convert_tag(node.tag):
-            text = convert_fn(node, text)
-
-        text += self.process_text(node.tail)
+        if not children_only:
+            convert_fn = getattr(self, 'convert_%s' % node.name, None)
+            if convert_fn and self.should_convert_tag(node.name):
+                text = convert_fn(node, text)
 
         return text
 
@@ -44,7 +76,7 @@ class MarkdownConverter(object):
         return escape(whitespace_re.sub(' ', text or ''))
 
     def __getattr__(self, attr):
-        # Handle heading levels > 2
+        # Handle headings
         m = convert_heading_re.match(attr)
         if m:
             n = int(m.group(1))
@@ -60,13 +92,18 @@ class MarkdownConverter(object):
 
     def should_convert_tag(self, tag):
         tag = tag.lower()
-        if self.tags_to_strip is not None:
-            return tag not in self.tags_to_strip
-        elif self.tags_to_convert is not None:
-            return tag in self.tags_to_convert
+        strip = self.options['strip']
+        convert = self.options['convert']
+        if strip is not None:
+            return tag not in strip
+        elif convert is not None:
+            return tag in convert
         else:
             return True
 
+    def indent(self, text, level):
+        return line_beginning_re.sub('\t' * level, text) if text else ''
+
     def underline(self, text, pad_char):
         text = (text or '').rstrip()
         return '%s\n%s\n\n' % (text, pad_char * len(text)) if text else ''
@@ -74,6 +111,9 @@ class MarkdownConverter(object):
     def convert_a(self, el, text):
         href = el.get('href')
         title = el.get('title')
+        if self.options['autolinks'] and text == href and not title:
+            # Shortcut syntax
+            return '<%s>' % href
         title_part = ' "%s"' % title.replace('"', r'\"') if title else ''
         return '[%s](%s%s)' % (text or '', href, title_part) if href else text or ''
 
@@ -89,24 +129,46 @@ class MarkdownConverter(object):
     def convert_em(self, el, text):
         return '*%s*' % text if text else ''
 
-    def convert_h1(self, el, text):
-        return self.underline(text, '=')
-
-    def convert_h2(self, el, text):
-        return self.underline(text, '-')
-
     def convert_hn(self, n, el, text):
-        return '%s %s\n\n' % ('#' * n, text.rstrip()) if text else ''
+        style = self.options['heading_style']
+        text = text.rstrip()
+        if style == UNDERLINED and n <= 2:
+            line = '=' if n == 1 else '-'
+            return self.underline(text, line)
+        hashes = '#' * n
+        if style == ATX_CLOSED:
+            return '%s %s %s\n\n' % (hashes, text, hashes)
+        return '%s %s\n\n' % (hashes, text)
 
     def convert_i(self, el, text):
         return self.convert_em(el, text)
 
+    def convert_list(self, el, text):
+        nested = False
+        while el:
+            if el.name == 'li':
+                nested = True
+                break
+            el = el.parent
+        if nested:
+            text = '\n' + self.indent(text, 1)
+        return text
+
+    convert_ul = convert_list
+    convert_ol = convert_list
+
     def convert_li(self, el, text):
-        parent = el.getparent()
-        if parent is not None and parent.tag == 'ol':
+        parent = el.parent
+        if parent is not None and parent.name == 'ol':
             bullet = '%s.' % (parent.index(el) + 1)
         else:
-            bullet = '*'
+            depth = -1
+            while el:
+                if el.name == 'ul':
+                    depth += 1
+                el = el.parent
+            bullets = self.options['bullets']
+            bullet = bullets[depth % len(bullets)]
         return '%s %s\n' % (bullet, text or '')
 
     def convert_p(self, el, text):
@@ -115,7 +177,13 @@ class MarkdownConverter(object):
     def convert_strong(self, el, text):
         return '**%s**' % text if text else ''
 
+    def convert_img(self, el, text):
+        alt = el.attrs.get('alt', None) or ''
+        src = el.attrs.get('src', None) or ''
+        title = el.attrs.get('title', None) or ''
+        title_part = ' "%s"' % title.replace('"', r'\"') if title else ''
+        return '![%s](%s%s)' % (alt, src, title_part)
 
-def markdownify(html, strip=None, convert=None):
-    converter = MarkdownConverter(strip, convert)
-    return converter.convert(html)
+
+def markdownify(html, **options):
+    return MarkdownConverter(**options).convert(html)

markdownify/pkgmeta.py

@@ -0,0 +1,8 @@
+pkgmeta = dict(
+    __title__='markdownify',
+    __author__='Matthew Tretter',
+    __version__='0.4.0',
+)
+
+globals().update(pkgmeta)
+__all__ = pkgmeta.keys()

markdownify/version.py

@@ -1 +0,0 @@
-__version__ = '0.1.0'

runtests.py

@@ -1,5 +0,0 @@
-#!/usr/bin/env python
-from nose.core import run, collector
-
-if __name__ == '__main__':
-    run()

setup.cfg

@@ -0,0 +1,2 @@
+[flake8]
+ignore = E501

setup.py

@@ -2,31 +2,80 @@
 import codecs
 import os
 from setuptools import setup, find_packages
+from setuptools.command.test import test as TestCommand, Command
 
 
 read = lambda filepath: codecs.open(filepath, 'r', 'utf-8').read()
-execfile(os.path.join(os.path.dirname(__file__), 'markdownify', 'version.py'))
+
+
+pkgmeta = {}
+execfile(os.path.join(os.path.dirname(__file__), 'markdownify', 'pkgmeta.py'),
+         pkgmeta)
+
+
+class PyTest(TestCommand):
+    def finalize_options(self):
+        TestCommand.finalize_options(self)
+        self.test_args = ['tests', '-s']
+        self.test_suite = True
+
+    def run_tests(self):
+        import pytest
+        errno = pytest.main(self.test_args)
+        raise SystemExit(errno)
+
+
+class LintCommand(Command):
+    """
+    A copy of flake8's Flake8Command
+
+    """
+    description = "Run flake8 on modules registered in setuptools"
+    user_options = []
+
+    def initialize_options(self):
+        pass
+
+    def finalize_options(self):
+        pass
+
+    def distribution_files(self):
+        if self.distribution.packages:
+            for package in self.distribution.packages:
+                yield package.replace(".", os.path.sep)
+
+        if self.distribution.py_modules:
+            for filename in self.distribution.py_modules:
+                yield "%s.py" % filename
+
+    def run(self):
+        from flake8.engine import get_style_guide
+        flake8_style = get_style_guide(config_file='setup.cfg')
+        paths = self.distribution_files()
+        report = flake8_style.check_files(paths)
+        raise SystemExit(report.total_errors > 0)
 
 
 setup(
-    name='python-markdownify',
+    name='markdownify',
     description='Convert HTML to markdown.',
     long_description=read(os.path.join(os.path.dirname(__file__), 'README.rst')),
-    version=__version__,
-    author='Matthew Tretter',
-    author_email='matthew@exanimo.com',
+    version=pkgmeta['__version__'],
+    author=pkgmeta['__author__'],
+    author_email='m@tthewwithanm.com',
     url='http://github.com/matthewwithanm/python-markdownify',
     download_url='http://github.com/matthewwithanm/python-markdownify/tarball/master',
     packages=find_packages(),
     zip_safe=False,
     include_package_data=True,
+    setup_requires=[
+        'flake8',
+    ],
     tests_require=[
-        'nose',
-        'unittest2',
+        'pytest',
     ],
     install_requires=[
-        'lxml',
-        'BeautifulSoup',
+        'beautifulsoup4',
     ],
     classifiers=[
         'Environment :: Web Environment',
@@ -39,6 +88,8 @@ setup(
         'Programming Language :: Python :: 2.7',
         'Topic :: Utilities'
     ],
-    setup_requires=[],
-    test_suite='runtests.collector',
+    cmdclass={
+        'test': PyTest,
+        'lint': LintCommand,
+    },
 )

tests.py

@@ -1,123 +0,0 @@
-import unittest
-from markdownify import markdownify as md
-
-
-class BasicTests(unittest.TestCase):
-
-    def test_single_tag(self):
-        self.assertEqual(md('<span>Hello</span>'), 'Hello')
-
-    def test_soup(self):
-        self.assertEqual(md('<div><span>Hello</div></span>'), 'Hello')
-
-    def test_whitespace(self):
-        self.assertEqual(md(' a  b \n\n c '), ' a b c ')
-
-
-class ArgTests(unittest.TestCase):
-
-    def test_strip(self):
-        self.assertEqual(
-            md('<a href="https://github.com/matthewwithanm">Some Text</a>', strip=['a']),
-            'Some Text')
-
-    def test_do_not_strip(self):
-        self.assertEqual(
-            md('<a href="https://github.com/matthewwithanm">Some Text</a>', strip=[]),
-            '[Some Text](https://github.com/matthewwithanm)')
-
-    def test_convert(self):
-        self.assertEqual(
-            md('<a href="https://github.com/matthewwithanm">Some Text</a>', convert=['a']),
-            '[Some Text](https://github.com/matthewwithanm)')
-
-    def test_do_not_convert(self):
-        self.assertEqual(
-            md('<a href="https://github.com/matthewwithanm">Some Text</a>', convert=[]),
-            'Some Text')
-
-
-class EscapeTests(unittest.TestCase):
-
-    def test_underscore(self):
-        self.assertEqual(md('_hey_dude_'), '\_hey\_dude\_')
-
-    def test_xml_entities(self):
-        self.assertEqual(md('&amp;'), '&')
-
-    def test_named_entities(self):
-        self.assertEqual(md('&raquo;'), u'\xbb')
-
-    def test_hexadecimal_entities(self):
-        # This looks to be a bug in BeautifulSoup (fixed in bs4) that we have to work around.
-        self.assertEqual(md('&#x27;'), '\x27')
-
-    def test_single_escaping_entities(self):
-        self.assertEqual(md('&amp;amp;'), '&amp;')
-
-
-class ConversionTests(unittest.TestCase):
-
-    def test_a(self):
-        self.assertEqual(
-            md('<a href="http://google.com">Google</a>'),
-            '[Google](http://google.com)'
-        )
-
-    def test_a_with_title(self):
-        self.assertEqual(
-            md('<a href="http://google.com" title="The &quot;Goog&quot;">Google</a>'),
-            r'[Google](http://google.com "The \"Goog\"")'
-        )
-
-    def test_b(self):
-        self.assertEqual(md('<b>Hello</b>'), '**Hello**')
-
-    def test_blockquote(self):
-        self.assertEqual(md('<blockquote>Hello</blockquote>').strip(), '> Hello')
-
-    def test_nested_blockquote(self):
-        self.assertEqual(
-            md('<blockquote>And she was like <blockquote>Hello</blockquote></blockquote>').strip(),
-            '> And she was like \n> > Hello'
-        )
-
-    def test_br(self):
-        self.assertEqual(md('a<br />b<br />c'), 'a  \nb  \nc')
-
-    def test_em(self):
-        self.assertEqual(md('<em>Hello</em>'), '*Hello*')
-
-    def test_h1(self):
-        self.assertEqual(md('<h1>Hello</h1>'), 'Hello\n=====\n\n')
-
-    def test_h2(self):
-        self.assertEqual(md('<h2>Hello</h2>'), 'Hello\n-----\n\n')
-
-    def test_hn(self):
-        self.assertEqual(md('<h3>Hello</h3>'), '### Hello\n\n')
-        self.assertEqual(md('<h6>Hello</h6>'), '###### Hello\n\n')
-
-    def test_i(self):
-        self.assertEqual(md('<i>Hello</i>'), '*Hello*')
-
-    def test_ol(self):
-        self.assertEqual(md('<ol><li>a</li><li>b</li></ol>'), '1. a\n2. b\n')
-
-    def test_p(self):
-        self.assertEqual(md('<p>hello</p>'), 'hello\n\n')
-
-    def test_strong(self):
-        self.assertEqual(md('<strong>Hello</strong>'), '**Hello**')
-
-    def test_ul(self):
-        self.assertEqual(md('<ul><li>a</li><li>b</li></ul>'), '* a\n* b\n')
-
-
-class AdvancedTests(unittest.TestCase):
-
-    def test_nested(self):
-        self.assertEqual(
-            md('<p>This is an <a href="http://example.com/">example link</a>.</p>'),
-            'This is an [example link](http://example.com/).\n\n'
-        )

tests/test_advanced.py

@@ -0,0 +1,6 @@
+from markdownify import markdownify as md
+
+
+def test_nested():
+    text = md('<p>This is an <a href="http://example.com/">example link</a>.</p>')
+    assert text == 'This is an [example link](http://example.com/).\n\n'

tests/test_args.py

@@ -0,0 +1,25 @@
+"""
+Test whitelisting/blacklisting of specific tags.
+
+"""
+from markdownify import markdownify as md
+
+
+def test_strip():
+    text = md('<a href="https://github.com/matthewwithanm">Some Text</a>', strip=['a'])
+    assert text == 'Some Text'
+
+
+def test_do_not_strip():
+    text = md('<a href="https://github.com/matthewwithanm">Some Text</a>', strip=[])
+    assert text == '[Some Text](https://github.com/matthewwithanm)'
+
+
+def test_convert():
+    text = md('<a href="https://github.com/matthewwithanm">Some Text</a>', convert=['a'])
+    assert text == '[Some Text](https://github.com/matthewwithanm)'
+
+
+def test_do_not_convert():
+    text = md('<a href="https://github.com/matthewwithanm">Some Text</a>', convert=[])
+    assert text == 'Some Text'

tests/test_basic.py

@@ -0,0 +1,13 @@
+from markdownify import markdownify as md
+
+
+def test_single_tag():
+    assert md('<span>Hello</span>') == 'Hello'
+
+
+def test_soup():
+    assert md('<div><span>Hello</div></span>') == 'Hello'
+
+
+def test_whitespace():
+    assert md(' a  b \n\n c ') == ' a b c '

tests/test_conversions.py

@@ -0,0 +1,122 @@
+from markdownify import markdownify as md, ATX, ATX_CLOSED
+import re
+
+
+nested_uls = re.sub('\s+', '', """
+    <ul>
+        <li>1
+            <ul>
+                <li>a
+                    <ul>
+                        <li>I</li>
+                        <li>II</li>
+                        <li>III</li>
+                    </ul>
+                </li>
+                <li>b</li>
+                <li>c</li>
+            </ul>
+        </li>
+        <li>2</li>
+        <li>3</li>
+    </ul>""")
+
+
+def test_a():
+    assert md('<a href="http://google.com">Google</a>') == '[Google](http://google.com)'
+
+
+def test_a_with_title():
+    text = md('<a href="http://google.com" title="The &quot;Goog&quot;">Google</a>')
+    assert text == r'[Google](http://google.com "The \"Goog\"")'
+
+
+def test_a_shortcut():
+    text = md('<a href="http://google.com">http://google.com</a>')
+    assert text == '<http://google.com>'
+
+
+def test_a_no_autolinks():
+    text = md('<a href="http://google.com">http://google.com</a>', autolinks=False)
+    assert text == '[http://google.com](http://google.com)'
+
+
+def test_b():
+    assert md('<b>Hello</b>') == '**Hello**'
+
+
+def test_blockquote():
+    assert md('<blockquote>Hello</blockquote>').strip() == '> Hello'
+
+
+def test_nested_blockquote():
+    text = md('<blockquote>And she was like <blockquote>Hello</blockquote></blockquote>').strip()
+    assert text == '> And she was like \n> > Hello'
+
+
+def test_br():
+    assert md('a<br />b<br />c') == 'a  \nb  \nc'
+
+
+def test_em():
+    assert md('<em>Hello</em>') == '*Hello*'
+
+
+def test_h1():
+    assert md('<h1>Hello</h1>') == 'Hello\n=====\n\n'
+
+
+def test_h2():
+    assert md('<h2>Hello</h2>') == 'Hello\n-----\n\n'
+
+
+def test_hn():
+    assert md('<h3>Hello</h3>') == '### Hello\n\n'
+    assert md('<h6>Hello</h6>') == '###### Hello\n\n'
+
+
+def test_atx_headings():
+    assert md('<h1>Hello</h1>', heading_style=ATX) == '# Hello\n\n'
+    assert md('<h2>Hello</h2>', heading_style=ATX) == '## Hello\n\n'
+
+
+def test_atx_closed_headings():
+    assert md('<h1>Hello</h1>', heading_style=ATX_CLOSED) == '# Hello #\n\n'
+    assert md('<h2>Hello</h2>', heading_style=ATX_CLOSED) == '## Hello ##\n\n'
+
+
+def test_i():
+    assert md('<i>Hello</i>') == '*Hello*'
+
+
+def test_ol():
+    assert md('<ol><li>a</li><li>b</li></ol>') == '1. a\n2. b\n'
+
+
+def test_p():
+    assert md('<p>hello</p>') == 'hello\n\n'
+
+
+def test_strong():
+    assert md('<strong>Hello</strong>') == '**Hello**'
+
+
+def test_ul():
+    assert md('<ul><li>a</li><li>b</li></ul>') == '* a\n* b\n'
+
+
+def test_nested_uls():
+    """
+    Nested ULs should alternate bullet characters.
+
+    """
+    assert md(nested_uls) == '* 1\n\t+ a\n\t\t- I\n\t\t- II\n\t\t- III\n\t\t\n\t+ b\n\t+ c\n\t\n* 2\n* 3\n'
+
+
+def test_bullets():
+    assert md(nested_uls, bullets='-') == '- 1\n\t- a\n\t\t- I\n\t\t- II\n\t\t- III\n\t\t\n\t- b\n\t- c\n\t\n- 2\n- 3\n'
+
+
+def test_img():
+    assert md('<img src="/path/to/img.jpg" alt="Alt text" title="Optional title" />') == '![Alt text](/path/to/img.jpg "Optional title")'
+    assert md('<img src="/path/to/img.jpg" alt="Alt text" />') == '![Alt text](/path/to/img.jpg)'

tests/test_escaping.py

@@ -0,0 +1,22 @@
+from markdownify import markdownify as md
+
+
+def test_underscore():
+    assert md('_hey_dude_') == '\_hey\_dude\_'
+
+
+def test_xml_entities():
+    assert md('&') == '&'
+
+
+def test_named_entities():
+    assert md('»') == u'\xbb'
+
+
+def test_hexadecimal_entities():
+    # This looks to be a bug in BeautifulSoup (fixed in bs4) that we have to work around.
+    assert md(''') == '\x27'
+
+
+def test_single_escaping_entities():
+    assert md('&') == '&'